Background In the recent years several predictors of radiographic spinal progression/syndesmophyte formation in axial spondyloarthritis (SpA) were identified: syndesmophytes at baseline (the strongest predictor so far), smoking, elevated level of the C-reactive protein. Furthermore, a number of predictive biomarkers (such as matrix metalloproteinase 3, procollagen II N-terminal propeptide, Wnt-antagonists) were identified, all, however, with a rather modest predictive value. Vascular endothelial growth factor (VEGF) is an essential mediator of the endochondral ossification and, therefore, might play a pathogenetic role in the process of syndesmophyte formation in axial SpA.

Objectives The aim of the study was to investigate the role of serum VEGF as a predictor of radiographic spinal progression in patients with axial SpA.

Methods Altogether 172 patients with definite axial SpA (95 with ankylosing spondylitis (AS) and 77 with non-radiographic axial SpA) from the German Spondyloarthritis Inception Cohort (GESPIC) were included in the current study. Spinal radiographs (lumbar and cervical spine, lateral views) were scored independently by two trained readers in a concealed and randomly selected order according to the modified Stoke Ankylosing Spondylitis Spine Score (mSASSS) scoring system. In addition to lateral views, anteroposterior views of the lumbar spine were scored for the presence of syndesmophytes. Radiographic spinal progression was defined as 1) mSASSS worsening by ≥2 units after 2 years, and 2) development of a new syndesmophyte or progression of existing syndesmophytes (i.e., formation of a bridging syndesmophyte) after 2 years. Serum VEGF levels were detected at baseline.

Results Mean baseline VEGF values were significantly higher in patients with mSASSS worsening by ≥2 units after 2 years (n=22) as compared to those without progression (562±357 vs 402±309 pg/ml, respectively, p=0.027) and in patients with syndesmophyte formation/progression (n=18) as compared again to those without new bone formation (579±386 vs 404±307 pg/ml, respectively, p=0.041). Area under the curve (AUC) was 0.646 (95% CI 0.52-0.77), p=0.027 for the mSASSS worsening ≥2 units and 0.648 (95% CI 0.51-0.79), p=0.041 for syndesmophyte formation. Importantly, the performance of VEGF as a predictor of radiographic spinal progression was better in patients who were already at high risk for such a progression due to the presence of syndesmophytes at baseline (n=48): AUC was 0.812 (95% CI 0.67-0.95), p=0.001, and 0.772 (95% CI 0.61-0.93), p=0.003, respectively. VEGF serum level of >600 pg/ml had a sensitivity of 53%, a specificity of 97%, and an odds ratio (OR) = 36.6 (95% CI 3.9-341.5) as a predictor of mSASSS worsening by ≥2 units over 2 years (after adjustment for elevated CRP and smoking OR was 37.2, 95% CI 3.5-392.8). The same serum level of VEGF demonstrated a sensitivity of 47%, a specificity of 94%, and an OR = 13.6 (95% CI 2.4-78.3) as a predictor of syndesmophyte formation (after adjustment for CRP and smoking OR = 14.0, 95% CI 2.3-85.6).

Conclusions High serum level of VEGF (>600 pg/ml) seems to be highly and independently predictive for radiographic spinal progression in patients with axial SpA especially in high-risk patients with already present syndesmophytes.

Disclosure of Interest None Declared